Photo-aza-Claisen rearrangements of cyclic enaminones - The

Grant R. Krow, Guoliang Lin, Keith P. Moore, Andrew M. Thomas, Charles DeBrosse, Charles W. Ross, III, and Harri G. Ramjit. Organic Letters 2004 6 (10...
0 downloads 0 Views 568KB Size
J. Org. Chem. 1989, 54, 4165-4168

4165

Photo-Aza-Claisen Rearrangements of Cyclic Enaminones B. Vogler, R. Bayer, M. Meller, and W. Kraus* Department of Chemistry, University of Hohenheim, 7000 Stuttgart 70, FRG

Fred M. &hell* Department of Chemistry, University of Tennessee, Knoxville, Tennessee 37996-1600 Received January 23, 1989

Cyclic enaminones 1-7 were prepared and photolyzed. Compounds 1-4 reacted as expected to provide azaDe Mayo products 10 and 11, whereas 5 provided enaminone 12 in a novel photo-aza-Claisen rearrangement. Compounds 6 and 7 ge.ve photo-aza-Claisen products 13 and 16, respectively, along with aza-De Mayo product 15 and 2 + 2 cycloaddition product 17, respectively. Photochemical reactions of enones with unactivated olefins leading to four-membered ring compounds are well known' and have often been used in natural products synthesis? Most of the known photochemistry deals with 2 + 2 photocycloaddition of a,@-unsaturatedketones or derivatives of l,&dicarbonyl compounds. Less attention has been paid to @-heteroatom-substituted compounds, especially where the heteroatom is a nitrogen, i.e., vinylogous amides (enaminones)or imides (N-acylenaminones) as the enone component. Typically, intermolecular photocyclization of simple enones to olefins provides 2 + 2 cycloaddition products often accompanied by compounds formally derived by ene3 reactions of the components. Despite a variety of intramolecular studies especially from the laboratories of Tamura? W i e ~ n e r ,Pattenden: ~ Oppolzer,2 Agosta,' and Scheffer,*it was not until 19769that a significant yield of an intramolecular ene product was reported. The first example of this reaction was discovered in the course of trying to obtain a simple 2 + 2 addition of a vinylogous amide, and similar intramolecular reactions using vinylogous imides'O and vinylogous estersll have now been reported. A principal stimulus for the investigation of @-heterosubstituted enones is due to the elegant work of De Mayo,lb who showed that initial addition products of vinylogous acids and esters could be transformed to 1,5-dicarbonyls via retro-aldol reaction. Likewise, others have shown that addition products of vinylogous amides can similarly be transformed into 1,5-keto imines12 via retro-Mannich (1)(a) Eaton, P. E. Acc. Chem. Res. 1968,I, 50. (b) De Mayo, P. Ibid. 1971,4,41.(c) Baldwin, S. W. In Organic Photochemistry; Padwa, A., Ed.; Marcel Dekker: New York, 1981;Vol. 5, p 123. (2)Oppolzer, W. Acc. Chem. Res. 1982,15, 135. (3)Recently an example was reported where the ene product is the main product. Berry, N. M.; Darey, M. C. P.; Harwood, L. M. Tetrahedron Lett. 1986,27,2319. (4)Tamura, Y.;Ishibashi, H.; Hirai, M.; Kita, Y.; Ikeda, M. J . Org. Chem. 1975,40,2702. (5)Wiesner, K.; Poon, L.; Jirkovsky, I.; Fishman, M. Can. J. Chem. 1969,47,433. (6)Birch, A. M.; Pattenden, G. J. Chem. SOC.,Perkin Trans. I 1983, 1913 and references cited therein. (7)Matlin, A. R.;Wolff, S.;Agosta, W. C. Tetrahedron Lett. 1983,24, 2961. Matlin, A. R.; George, C. F.; Wolff, S.; Agosta, W. C. J.Am. Chem. SOC. 1986,108,3385 and references cited therein. (8)Scheffer, J. R.; Wostradowski, R. A. J . Org. Chem. 1972,37,4317. (9)Tamura, Y.; Ishibashi, H.; Ikeda, M. J. Org. Chem. 1976,41,1277. (10)(a) Schell, F. M.; Cook, P. M. J. O g . Chem. 1978,43,4420. (b) Schell, F. M.; Cook, P. M.; Hawkinson, S. W.; Cassady, R. E.; Thiessen, W. E. Ibid. 1979,44, 1380. (11)Oppolzer, W.; Bird, T. G. C. Helu. Chim. Acta 1979,62,1199. (12)(a) Schell, F. M.; Cook, P. M. J. Org. Chem. 1984,49,4067.(b) For another example of intramolecular cycloaddition of a vinylogous amide, see: Winkler, J. D.; Hershberger, P. M.; Springer, J. P. Tetrahedron Lett. 1986,27,5177.

0022-3263/89/1954-4165$01.50/0

chemistry. The high yield of this reaction combined with the facile differentiation of the carbonyls produced in the retro-Mannich process made further study appealing. From the broad studies by Agosta7J3in the case of enones it is known that the outcome of the photochemical reaction is strongly structure dependent. Even simple structural changes such as introduction of a methyl group produced significant changes in the product ratio of 2 + 2 versus ene or "crossed" versus "straight" addition. In the case of vinylogous amides less systematic studies have been reported; however, simple 2 + 2 cycloaddition, a photo-ene reaction, and an example of the aza-De Mayo sequence have been reported. It was in this latter context that we investigated the photochemistry of compounds 1-7, only to discover in the case of 5 yet another reaction course for these unpredictable materials.

H

H

3

1 R:H 2 R=CH,

R=H

4 R=CH3

H 5

R 6 R:H 7 R:CH,

Starting with the known transformation of 8 to 9, we investigated the influence of structural changes including

8

9

substitution pattern, ring size variation, and olefin separation. Irradiation of compounds 1-4 in either methylene chloride or acetonitrile yielded the expected aza-De Mayo products loa, l o b , l l a , and l l b , respectively, in even better yields than previously reported. When the linking unit was shortened, a change in the direction of addition was expected, as reported earlier by Tamura.4 Irradiation of 5 in acetonitrile produced a single characterizable compound along with starting material and (13)Wolff, S.;Agosta, W. C. J.Am. Chem. SOC.1983,105,1292,1299.

0 1989 American Chemical Society

4166 J. Org. Chem., Vol. 54, No. 17, 1989

Vogler et al.

some polymeric material. The yield of the new product was 57% on the basis of consumed 5. Inspection of the 13C NMR spectrum surprisingly revealed the product to be an olefinic enone. Furthermore, the enone and one olefinic carbon were nonprotonated and the remaining olefinic carbon was a methylene. The proton NMR corroborated these observations with olefinic absorptions a t 4.93 (1 H) and 4.67 (1 H) and also a broad resonance at 4.97 (2 H) which proved to have a very concentrationdependent chemical shift. These data and the presence of a single aliphatic methine suggested structure 12 for the product formed via a photo-aza-Claisen rearrangement. This is fully supported by detailed analysis of homo- and heteronuclear 2-D COSY spectra and 2-D COLOC spectra, which clearly define the connection between the methine carbon and both olefinic systems.

Comparison of the cyclohexane methylene chemical shifts in compounds 9 and 15 shows them to be quite dissimilar. In the former case the six-membered ring has been assumed to be in the chair conformation, having diaxial substitution to account for the high methylene values, but this cannot be the situation for 15. Furthermore, if the amino methylene-hydrogen relationship in 14 is cis, then there would have to be a serious axial interaction of the amino methylene with a chair cyclohexane ring. Thus, either the cyclohexane ring cannot be in the chair conformation or the relationship of the hydrogen and amino methylene must be trans. Irradiation of one of the amino methylene hydrogens does produce an enhancement in the proton absorption at the nearby keto methylene which is in accord with the trans relationship and a rigid cyclohexane ring. Introducing a methyl group a t the nitrogen atom in 6 produced 7, which upon irradiation provided two products that could be isolated and characterized, although in low yield. One of them is the photuaza-Claisen rearrangement product 16, which was easily recognized by comparison of

16

12

In order to find out whether this transformation is a singular example or a synthetically useful reaction, we have also examined structurally related systems. Changing the ring size in the olefin part from a five-membered to a six-membered ring led to compound 61°bwhose irradiation provided two products. One of them can be rationalized by its NMR spectra as the expected photo-aza-Claisen rearrangement product 13. The second is a compound n

13

14

15

whose IR and 13C NMR spectra quickly reveal its imino ketone nature and thus its De Mayo origins. The nature of the imino carbon with a very high chemical shift (198.2 ppm) was only recognized with certainty due to the IR stretch at 1640 cm-’. The latter value is also abnormal for imines in these systems where we and others have rep ~ r t e d ’ ~ J *values J ~ in the range 1610-1625 cm-’. The abnormalities in these values forced us to consider the “straight adduct” 15, whereas production of “crossed adduct” 14 would have seemed more likely. In fact, close inspection of the 13C NMR data demonstrated that 15 is, indeed, the “straight” addition product. Thus, the quaternary center in the six-membered ring has a low-field shift as in 9 while the quaternary in the “crossed adduct” 14 would be expected to appear a t much higher field.I6 Likewise, the methine center in 15 appears a t 36.3 ppm (37.6 ppm in 9), which is higher than would be expected if next to an imino carbonyl center. The difficult part of the structure assignment is the stereochemicalrelationship of the methine hydrogen and the amino methylene. (14) Swindell, C. S.;Patel, B. P.; deSolms, S. J.; Springer, J. P. J. Org. Chem. 1987,52, 2346. (15) See Experimental Section. (16) See, for example, the quaternary center of the crossed adduct in Swindell e t al., ref 14.

17

spectral details with those from compound 13 above. The other product was the expected “crossedn cycloaddition product whose structure was partially defined as 17 by NMR analysis.” Thus, the keto methine appears as a broad singlet, which was shown to be W-coupled to one of the amino methylene protons by using a COSY analysis. There are four possible stereo structures for this adduct. Only those two in which the keto methine and the amino methylene bridge are anti could display the observed W-coupling. The photo-aza-Claisen products reported herein can be rationalized by assuming the intermediacy of a diradical such as 18, which can unravel directly to the keto imine tautomer 19. However, we have no evidence to rule out

18

19

20

direct formation of, or collapse of the diradical 18 to, [2.2.0]bicycle 20 followed by cycloreversion to the same tautomer. This possible “straight” addition of 1,Bhexadienes has now been observed with vinylogous amides, for example, as reported above in the conversion of 6 to 15. The net reaction, 5 to 12, is a photo-aza-Claisen rearrangement,18J9 which to our knowledge is the first example of such a photochemically induced process involving (17) The cyclohexane-cyclobutane ring fusion in 17 is most likely cis on the basis of comparison of chemical shift data with that reported“ for a similar photoproduct with an N-acetyl rather than N-methyl substituent. Except for those carbons directly affected by the change in nitrogen substitution, chemical shifts agree well. Unfortunately, this congruence is blemished by a disagreement in multiplicity aasignmenta,in particular, at least one of the signals reported for the neopentyl methyls in the reference compound is at too high a field for such substitution. (18) Lutz, R. P.Chem. Reu. 1984,84, 205. For certainty the starting material w a also ~ heated to reflux in benzene, whereas the photochemistry was performed at 16 O C . No thermal rearrangement was detectable. (19) An analogous photo-Claisen rearrangement is reported by Smith, A. B., 111; Agosta, W. C. J. Am. Chem. SOC.1973, 95, 1961.

Cyclic Enaminone Photo-Aza-Claisen Rearrangements

J. Org. Chem., Vol. 54, No. 17, 1989 4167

ethyl)amine;20reaction time 18 h; yield 13.7 g (67%) of 3; mp 114-116 OC (acetone/diethyl ether). 'H N M R 5.44 (m, 1 H), 5.13 (8, 1H), 4.77 (bs,1H), 3.18 (q,2 H), 2.1-2.45 (m, 10 H), 1.8-2.0 (m, 4 H). 13CNMR CH2 (40.60, 36.13,34.50, 32.11,29.45, 29.14, 22.89, 21.64), CH (125.57, 114.54), C (197.01, 165.2, 140.7). MS: calcd for ClJ-I1fiO 205.1467, found 205.1467; fragments (intensity) Experimental Section 205 (38), 125 (84), 124 (loo), 112 (40), 96 (34). IR (CHC13)cm-': 3400 (N-H), 3260 (NH), 2840 (CH), 1580 (vin. amide), 1540 (vin. NMR spectroscopyemployed a Bruker WM 250 spectrometer. amide). Proton spectra were obtained at 250 MHz with CDC13as solvent 3-[(2-Cyclopentenylethyl)amino]-5,5-dimethyl-2-cycloand tetramethylsilane as internal standard. Carbon-13 spectra hexen-1-one (4): 14.0 g of 5,5-dimethyl-1,3-cyclohexanedione, were obtained operating at 62.89 MHz with CDC13as solvent and 12.4 g of (2-cyclopentenylethy1)amine;reaction time 17 h; yield standard (CDC13 = 77.00 ppm). Multiplicities of carbons were 15.6 g (67%) of 4; mp 110-113 OC (diethyl ether). 'H NMR: 5.44 determined by use of DEFT experiments. Infrared spectroscopy (m, 1 H), 5.11 (s, 1H), 4.78 (bs, 1H), 3.19 (ddd, 2 H), 2.4-2.2 (m, employed a Zeiss IMR-25 spectrometer, using CHC13 or CC14 6 H), 2.18 (s, 4 H), 1.87 (m, 2 H), 1.06 ( 8 , 6 H). 13CNMR: CH3 solutions, films on NaCl plates, or KBr pellets. Mass spectroscopy (27.87), CH2 (50.03,40.71, 34.51, 32.35, 32.11, 29,47, 22.90), CH employed a Varian MAT 311 A (high resolution) 44 S (capillary (125.56,94.37),C (196.42,163.89,140.69,32.17).MS: calcd for GC-MS) instrument at 70 eV. Photochemical reactions were run Cl5HZ3NO233.1779, found 233.1779; fragments (intensity) 233 with use of a low pressure mercury lamp, HANAU 5OC-750 W, (20), 153 (54), 152 (48), 140 (36), 138 (40), 94 (20), 83 (100). IR in a Pyrex glass apparatus. (KBr) cm-': 3220 (NH), 3010 (C=CH), 2940 (CH), 2860 (CH), Preparation of Amines. A suspension of LAH in dry diethyl 1580 (vin. amide), 1560 (vin. amide), 1530 (vin. amide). ether was placed in a three-necked flask equipped with a reflux 3 4 (Cyclopentenylmethyl)amino]-5,5-dimet hyl-2-cyclocondenser, a drying tube, and an ice bath. The nitrile dissolved hexen-1-one (5): 1.2 g of (cyclopentenylmethyl)amine,1.9 g of in dry ether was dropped slowly into the reaction flask by means 5,5-dimethyl-l,3-cyclohexanedione; reaction time 16 h; yield 1.5 of a dropping funnel. The color of the reaction mixture turned g (57.0%) of 5,mp 139 "C (petroleum ether/acetone). 'H NMR to yellow green. After the addition was finished, the cooling bath 5.55 (m, 1H), 5.22 (bs, 1H), 5.1 (s, 1H), 3.75 (m, 2 H), 2.26-2.37 was removed and the resulting solution stirred for another 14 h (m, 4 H), 2.22 (s, 2 H), 2.18 (s, 2 H), 1.84-1.96 (m, 2 H), 1.07 (s, at room temperature and 1 h at reflux. Hydrolysis was accom6 H). '3C NMR: CHS (28.07), CH2 (50.13,43.35,43.08,33.27,32.11, plished by addition of the stoichiometric amount of water under 23.11), CH (126.49,94.94), C (196.49, 163.62, 139.17,32.56). MS: external cooling with ice. The solids were filtered and washed calcd for Cl4HZ1NO219.1618, found 219.1624; fragments (intensity) well with ether. The ethereal solution was dried over anhydrous 219 (W), 204 (70), 191 (80), 190 (561,163 (20), 135 (100). IR (KBr) magnesium sulfate and the solvent was removed on a rotatory 3050 (C=CH), 2960 (CH), 1595 (vin. cm-': 3420 (NH), 3220 (NH), evaporator. The crude amine then was purified by short path amide), 1560 (vin. amide), 1540 (vin. amide). distillation to prevent loss of the fairly volatile material. 3 4(Cyclo hexenylmet hyl)methylamino]-5,5-dimethyl-2(Cyclopentenylmethy1)amine: 5.09 g of 1-cyano-1-cyclocyclohexen-1-one (7): 2.8 g of 3-[(cyclohexenylmethyl)pentene, 4.68 g of LiAlH4,10.17 g of water, 400 mL of diethyl ether; amino]-5,5-dimethyl-2-cyclohexen-l-one (6),lob960 mg of NaH yield 2.6 g (49%); bp (15 Torr) = 62 "C. (60% suspension in oil), 1.5 mL of methyl iodide, 150 mL of dry (Cyclohexenylmethy1)amine: 16.4 g of 1-cyano-1-cyclodioxane. The vinylogous amide 6 and NaH were placed in dry hexene, 12.73 g of LiAlH4, 24.12 g of water, 350 mL of diethyl dioxane, refluxed for 1h, and then cooled to room temperature. ether; yield 12.1 g (73%); bp (15 Torr) = 55 OC. Methyl iodide then was added by means of a dropping funnel. General Procedure for t h e Preparation of Vinylogous After the addition was complete the solution was refluxed again Amides. The amine and the appropriate dione were dissolved for 2 h. The solution was concentrated under vacuum and the in benzene and heated to reflux with continuous removal of water residue dissolved in methylene chloride whereupon a white by a Dean-Stark trap. After water removal ceased, benzene was precipitate formed. The solid was removed by filtration, the distilled from the reaction mixture. The residual orange oil was solution dried over magnesium sulfate, and the methylene chloride dissolved in methylene chloride and washed twice with 5% NaOH removed to provide an orange-brown crystalline compound. solution (50 mL) and once with water. After drying over magRecrystallization from diethyl ether yielded 1.66 g (56%) of 7, nesium sulfate, methylene chloride was removed by distillation mp 96 "C. 'H N M R 5.40 (m, 1 H), 5.16 (s, 1 H), 3.73 (s, 2 H), and the remaining oil was chromatographed first on alumina 2.89 (8, 3 H), 2.27 (s, 2 H), 2.16 ( 8 , 2 H), 1.92-2.03 (m, 2 H), (Woelm) with chloroform and then on silica using methylene 1.69-1.85 (m, 2 H), 1.55-1.67 (m, 4 H), 1.06 (s, 6 H). 13C NMR: chloride/ethanol. The individual products were recrystallized CH3 (28.57, 38.16), CH2 (22.22, 22.31, 24.78, 26.01,40.31,49.30, from the solvents indicated below. 57.39), CH (97.32, 122.98), C (32.77, 132.00, 194.05, 196.24). MS: 3 4(2-Cyclohexenylethyl)amino]-2-cyclohexen-l-one (1): calcd for CIBH&JO247.1932, found 247.1928; IR (KBr) cm-': 2940 5.0 g of 1,3-cyclohexanedione, 6.1 g of (2-cyclohexenylethy1)(CH), 1600 (vin. amide), 1550 (vin. amide), 1500 (vin. amide). amine;lh reaction time 12 h; yield 7.3 g (75%) of 1; mp 106-107 Photochemistry of Vinylogous Amides. After dissolving "C (diethyl ether). 'H NMR 5.50 (m, 1 H), 5.43 (bs, 1 H), 5.10 the vinylogous amide in 600 mL of freshly distilled solvent (9, 1 H), 3.15 (4, 2 H), 2.35 (t, 2 H), 2.3 (t, 2 H), 2.23 (t, 2 H), (acetonitrile, unless otherwise stated), the solution was degassed 1.85-2.05 (m, 6 H), 1.5-1.7 (m, 4 H). 13CNMR: CH2 (40.52,36.43, with oxygen-free argon for 0.5 h and then photolyzed until TLC 36.43, 29.56, 27.84, 25.12, 22.67, 22.18, 21.95), CH (123.69, 96.23), indicated no further reaction. After removal of the solvent the C (196.98, 164.81, 134.14). MS: calcd for Cl4HZ1NO219.1623, residue was chromatographed on silica with solvent systems as found 219.1623;fragments (intensity) 219 (48), 125 (loo), 124 (96), indicated below. 112 (26), 96 (34), 67 (32). IR (CCq) cm-': 3420 (NH), 3260 (NH), Photolysis of 3-[(2-cyclohexenylethyl)amino]-2-cyclo2940 (CH), 1630 (vin. amide), 1590 (vin. amide), 1500 (vin.amide). 3 4 (2-Cyclohexenylethyl)amino]-6,6-dimethyl-2-cyclo- hexen-1-one (1): 2.5 g of 1, solvent acetonitrile; reaction time hexen-1-one (2): 12.5 g of 6,6-dimethyl-1,3-cyclohexanedione, 8 h; workup by flash chromatography on silica; methylene chloride/ethyl acetate 3:2 as solvent; yield 1.75 g (70%)of loa. 'H 12.3 g of (2-cyclohexenylethyl)amine;10*reaction time 14 h; yield NMR: 3.65 (m, 2 H), 3.3 (dd, 1 H), 2.6-2.0 (m, 9 H), 1.97 (dd, 17.4 g (79%) of 2; mp 83-85 "C (diethyl ether). 'H NMR: 5.50 1 H), 1.9-1.2 (m, 8 H). 13CNMR CH2 (56.20,43.76,43.76, 37.90, (m, 1 H), 4.96 (s, 1 H), 4.58 (bs, 1 H), 3.10 (ddd, 2 H), 2.40 (t,2 32.09, 30.17, 29.97, 26.23, 22.98, 20.94), CH (37.25), C (215.00, H), 2.23 (dd, 2 H), 1.9-2.1 (m, 4 H), 1.83 (t, 2 H), 1.5-1.7 (m, 4 183.02,56.13). MS: calcd for Cl4HZ1NO219.1623, found 219.1624; H), 1.1(s, 6 H). 13CNMR: CH3 (24.87), CH2 (40.33,40.33,36.34, fragments (intensity) 219 (loo), 191 (16), 162 (20), 150 (22), 124 35.57, 27.60, 26.21, 22.45, 21.97), CH (123.64, 94.81), C (202.23, (32), 109 (go), 108 (74). IR (KBr) cm-': 2960 (CH), 2875 (CH), 162.92, 134.19, 39.21). MS: calcd for C16H2,N0 247.1936, found 1715 (C=O), 1630 (C=N). 247.1937; fragments (intensity) 247 (50),153 (loo), 152 (40), 140 (30), 108 (40),96 (60). IR (CHCl,) cm-': 3400 (NH), 3280 (NH), 2860 (CH), 1580 (vin. amide), 1540 (vin. amide). 3 4 (2-Cyclopentenylethy1)amino]-2-cyclohexen-l-one (3): (20) Hanck, M.; Schneider, H.-J. Justus Liebigs Ann. Chem. 1965,686, 11.2 g of 1,3-~yclohexanedione,12.4 g of (2-cyclopentenyl16.

a nitrogen chromophore. The presence of a @-nitrogen substituent clearly provides enones with a tendency for the unusual and it will be of interest to see if they undergo other unexpected reactions.

4168

J. Org. Chem., Vol. 54, No. 17, 1989

Photolysis of 3-[(2-cyclohexenylethyl)amino]-6,6-dimethyl-2-cyclohexen-1-one (2): 400 mg of 2,solvent acetonitrile; reaction time 14 h; workup by chromatography on silica; methylene chloride/ethyl acetate 3:2 as solvent; yield 240 mg (60%) of lob; mp 103-105 "C (diethyl ether). 'H NMR 3.7-3.4 (m, 3 H), 2.85 (m, 1H), 2.45 (m, 2 H), 2.2-2.0 (m, 3 H), 1.9 (ddd, 1 H), 1.8-1.6 (m, 6 H), 1.4-1.3 (m, 3 H), 1.13 (s, 3 H), 1.11(8, 3 H). '3C NMR: CH3 (29.28, 20.53), CH2 (55.33,38.92,38.02,36.21,31.36, 30.05, 26.05, 21.94, 20.85), CH (36.11), C (216.71, 182.18,55.56, 47.10). MS: calcd for Cl6HZ5NO247.1947, found 247.1942; fragments (intensity) 247 (62), 219 (lo), 204 (36), 191 (64), 176 (18), 154 (64), 149 (56), 43 (100). IR (KBr) cm-': 2920 (CH), 2860 (CH), 1690 (C=O), 1610 (C=N). Photolysis of 3 4 (2-Cyclopentenylethyl)amino]-2-cyclohexen-1-one (3): 3.0 g of 3, solvent acetonitrile; reaction time 8 hhworkup by flash chromatography on silica; methylene chloride/ethyl acetate 3:2 as the solvent; yield 2.25 g (75%) of lla. 'H NMR 3.8-3.7 (ddd, 1H), 3.6-3.45 (ddd, 1H), 2.86 (dd, 1H), 2.6-1.4 (m, 16 H). '% NMR CH2(56.29,45.59,43.44,41.55,33.87, 32.02, 29.69,26.25,22.70), CH (44.77), C (214.22, 181.68,62.59). MS: calcd for C13HlBN0205.1461, found 205.1464; fragments (intensity) 205 (40), 177 (12), 162 (8), 150 (15), 148 (161,135 (16), 124 (24), 109 (32), 96 (46), 95 (loo),94 (74). IR (KBr) cm-': 2960 (CH), 2880 (CH), 1715 (C=O), 1635 (C=N). Photolysis of 3-[(2-cyclopentenylethyl)amino]-5,5-dimethyl-2-cyclohexen-1-one (4): 2.7 g of 4, solvent acetonitrile; reaction time 8 h; workup by flash chromatography on silica; methylene chloride/ethyl acetate 1:l as the solvent; yield 2.16 g (80%)of l l b ; mp 65-68 OC (diethyl ether). 'H NMR: 3.70 (ddd, 1 H), 3.55 (ddd, 1 H), 2.6 (dd, 1 H), 2.43 (AB, 1 H), 2.37 (AB, 1 H), 2.20 (AB, dd, m, 3 H), 1.90 (AB, 1 H), 1.70-1.85 (ddd, 1H), 1.3-1.7 (m, 7 H), 1.00 (s, 3 H), 0.90 (s,3 H). '% NMR: CH3 (29.60, 28.38), CH2 (56.02, 51.78, 46.84, 40.08, 39.13, 33.37, 31.26,19.94), CH (45.50), C (211.44, 177.88, 62.53, 34.08). MS: calcd for Cl5HZ3NO233.1774, found 233.1777; fragments (intensity) 233 (32), 218 (12), 205 (lo), 192 (14), 178 (20), 150 (34), 136 (24), 135 (29), 109 (72), 96 (100). IR (KBr) cm-': 2950 (CH), 2880 (CH), 2860 (CH), 1690 (C=O), 1615 (C=N). Photolysis of 3-[(cyclopentenylmethyl)amino]-5,5-dimethyl-2-cyclohexen-1-one (5): 800 mg of 5, solvent acetonitrile; reaction time 14.5 h; yield 280 mg of 5, 320 mg of 12;separation was accomplished by flash chromatography; solvent system methylene chloride/ethanol91. 12: oil. 'H N M R 4.97 (bs, 1 H),4.92 (m, 1H),4.66 (m, 1H),4.06 (m, 1H), 2.41(m,2 H), 2.27 ( 8 , 2 H), 2.23 (dd, 2 H), 1.78-1.85 (m, 2 H), 1.58-1.67 (m, 2 H), 1.08 (s,3 H), 1.04 (5, 3 H). 13C NMR: CH3 (27.69, 28.27), CH2 (104.95, 50.21, 44.0, 32.65, 29.92, 24.381, CH (39.47), C (195.15, 158.44,151.57,108.4,31.86). MS: calcd for Cl4HZlNO219.1618, found 219.1621; fragments (intensity) 219 (loo),204 (16), 190 (22), 176 (44), 162 (22), 140 (42), 135 (90). IR (KBr) cm-': 3340 (NH), 3200 (NH), 3080 (R2C=CH,), 2960 (CH), 2880 (CHI, 1650 (vin. amide), 1610 (vin. amide), 1550 (vin. amide). Photolysis of 3-[(cyclohexenylmethyl)amino]-5,5-dimethyl-2-cyclohexen-1-one(6): 1.0 g of 6;'Ob solvent acetonitrile;

Vogler et al. reaction time 18 h. After filtration over alumina with diethyl ether/ethanol(18:1), the resulting oily material was chromatographed under medium pressure on silica with ligroin/methylene chloride/ethanol(l621). The resulting two fractions were further purified by HPLC using the same solvent system: yield 30 mg of 13,35 mg of 15,174 mg of starting material, 40 mg of unresolved mixture. 13: oil. 'H NMR: 4.74 (m, 1 H), 4.55 (bs, 2 H), 4.45 (m, 1 H), 3.92 (m, 1 H), 2.20-2.50 (m, 6 H), 1.2-1.9 (m, 6 H), 1.07 (9, 3 H), 1.02 ( ~ , H). 3 13CNMR. CH3 (28.67, 27.54), CH2 (107.44, 50.15, 44.31, 35.82, 29.63, 27.45, 26.60), CH (37.17), C (195.02, 157.98,146.68, 111.00,31.88). MS: calcd for C15HBN0233.1780, found 233.1777; fragments (intensity) 233 (72),218 (loo),216 (20), 204 (30), 201 (32), 191 (18), 190 (32), 152 (24),83 (52). IR (CHCl3) cm-': 3360 (NH), 3240 (NH), 3100 (&C=CH2), 2960 (CH), 2880 (CH), 1715 (vin.amide), 1650 (vin.amide), 1620 (vin. amide), 1560 (vin. amide). 15 oil. 'H NMR 3.76 (d, 1 H), 3.38 (d, 1H), 2.75 (d, 1 H), 2.38 (dd, 1 H), 2.25 (dd, 1 H), 2.21 (m, 1 H), 2.16 (m, 1 H), 2.11 (m, 1 H), 2.00 (dd, 1 H), 1.55-1.77 (m, 8 H), 1.19 (e, N M R CH3 (33.60, 26.01), CH, (58.54, 3 H), 1.11 (s, 3 H). 50.23,49.50,42.67,34.20,30.16,25.73,22.92),CH (36.27),C (209.95, 198.24,56.01, 36.91). M S calcd for C15HzsN0233.1780, found 233.1777; fragments (intensity) 233 (14), 218 (14), 205 (14), 204 (14), 190 (20), 177 (20), 176 (26), 148 (34), 124 (42), 109 (loo), 94 (94). IR (CHC13)cm-': 2930 (CH), 2860 (CH), 1700 (C=O), 1640 (C=N). Photolysis of 3-[ (cyclohexenylmethyl)methylamino]-5,5dimethyl-2-cyclohexen-1-one (7): 600 mg of 7, solvent cyclohexaneltert-butyl alcohol (6:l); reaction time 10 h. After crude separation by flash chromatography on silica with methylene chloride/ethyl acetate (7:3), the nonpolar fraction was chromatographed with a medium pressure pump and a gradient solvent system with methylene chloride/ethyl acetate (955 changing to 7030): yield 56 mg of 17,lO mg of 16,409 mg of starting material. 16: oil. 'H NMR: 4.86 (m, 1 H), 4.71 (m, 1 H), 4.39 (m, 1 H), 3.95 (m, 1 H), 2.91 (d, 3 H), 2.39 (m, 1 H), 2.37 (AB, 2 H), 2.32 (s, 2 H), 2.15 (m, 1 H), 1.74-1.84 (m, 2 H), 1.50-1.70 (m, 3 H), 1.10 (9, 3 H), 1.06 (5, 3 H). 13C NMR CH3 (29.85, 29.01, 27.97), CH, (107.55, 49.86, 39.46, 35.77, 29.69, 27.44, 26.63), CH (37.12), MS: calcd for CI6H,NO C (193.48,160.17,147.04,107.55,31.48). 247.1936, found 247.1934; fragments (intensity) 247 (50),232 (loo), 218 (24),204 (40), 201 (22), 190 (20). IR (CHCld cm-': 3400 (NH), 3340 (NH), 3080 (R2C=CH2), 2940 (CH), 2860 (CH), 1705 (vin. amide), 1650 (vin. amide). 17: oil. 'H NMR: 3.30 (d, 1H), 2.78 (bs, 1 H), 2.26 (s,3 H), 2.0-2.19 (AB, 4 H), 1.91 (d, 1 H), 1.5-1.9 (m, 4 H), 1.5-1.3 (m, 3 H), 1.30 (9, 3 H), 1.1 (s, 3 H). 13CNMR: CH3 (36.66, 34.11,31.30), CH, (58.94,53.61, 34.97, 26.16, 23.88, 22.55, 22.02), CH (52.16, 45.02), C (209.84, 74.58, 49.32, 38.07). MS: calcd for C16H25N0247.1936, found 247.1932. The compound decomposes very rapidly.

Acknowledgment. We thank Studienstiftung des Deutschen Volkes (B.V.), Fonds der Chemischen Industrie (W.K., R.B., M.M.), and NATO grant No. 0272 (W.K., F.M.S.) for support of this investigation.